Meta CEO Mark Zuckerberg recently addressed various strategic choices facing the tech giant as it expands its footprint in artificial intelligence through the construction of extensive data centers. Speaking at the Meta Connect event held on September 25, 2024, at the company's headquarters in Menlo Park, California, Zuckerberg indicated that the company finds itself weighing whether to capitalize on its excess capacity by selling it off or to retain it for future use.
Unlike its three major competitors in the U.S. hyperscaler market, Meta has not ventured into the cloud services business, despite investing heavily in infrastructure. However, in light of recent developments, Zuckerberg has hinted at the prospect of launching a cloud service to leverage the company’s capacity as competition heats up. There have already been preliminary discussions, including one with the AI company Anthropic regarding leasing computing resources from Meta. "We're receiving numerous offers for compute at significantly higher rates than we paid," Zuckerberg commented during the company's recent earnings report, where he also mentioned that new coding and productivity tools are on the horizon.
Despite these opportunities, the atmosphere at Meta is turbulent. The company provided a lower-than-anticipated revenue outlook for the third quarter, with free cash flow plummeting 90% year-on-year due to increased capital expenditures. Following the announcement, Meta's stock took a hit, dropping over 7% in after-hours trading and marking an 11% decline since the start of the year.
In addition, Meta adjusted its capital expenditure projections for 2026, increasing the lower end by $5 billion, bringing the total forecast to between $130 billion and $145 billion. For context, Alphabet recently raised its forecast to $205 billion, and Microsoft indicated a capex range of around $175 billion. As Meta gears up for a more competitive landscape, investor interest is piqued regarding Zuckerberg's strategy in AI, an area where Meta trails behind leaders like OpenAI and Google.
While Zuckerberg's comments offered few concrete plans, he elaborated on the complex dynamics the company must navigate. "We need to make decisions on monetizing resources today versus investing in future capabilities," he explained. He emphasized the potential of not only selling capacity but also introducing API and productivity tools, as well as AI agents the company is developing. Notably, as Meta pursues its AI initiatives led by chief AI officer Alexandr Wang, the firm will need sufficient compute resources to drive its ambitions forward.
Zuckerberg acknowledged the challenges ahead, particularly in breaking into the enterprise sector where Meta has historically faced hurdles. He indicated that building a capable sales team would be vital for successful business engagement. The recent recruitment of Dave Brown, a former senior executive at Amazon Web Services, signals the seriousness of this transition.
Although Zuckerberg’s track record includes some missteps, such as the costly metaverse initiative—which continues to drain resources—he remains committed to diversifying Meta’s revenue sources, which still heavily rely on digital advertising. "I recognize that this is a significant industry-wide gamble," he concluded, expressing optimism that those who invest in Meta's AI developments will ultimately see rewards.
